Self-tuning Information Fusion Wiener Deconvolution Filter
Zili Deng · Science Technology and Engineering · 2006
For the multisensor signal deconvolution systems with unknown model parameter and noise statistics, by the modern time series analysis method, based on the on-line identification of the autoregressive moving average (ARMA) innovation model parameters, the noise variances can on-line be estimated, and a self-tuning information fusion Wiener deconvolution filter is presented. Its asymptotic optimality is proved, i.e. if the parameter estimation of the ARMA innovation model is consistent, then it will converges to the optimal fusion Wiener deconvolution filter with known noise statistics. Compared with the single-sensor case, its accuracy is improved. A simulation example for a deconvolution system with 3-sensor shows its effectiveness.